IRC log for #edev on 20070414

00:05.32*** join/#edev chouimat|away (n=dieu@r2351064.cidc.net)
03:15.13*** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net)
05:32.09*** join/#edev redblue (i=star@ppp098.108-253-207.mtl.mt.videotron.ca)
05:36.58*** part/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca)
06:56.00*** join/#edev DrZimmerman (n=theo@84-74-134-243.dclient.hispeed.ch)
08:39.13*** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-31-110.westend.de)
10:51.58*** join/#edev d-d-c (n=none@host48-143-dynamic.54-82-r.retail.telecomitalia.it)
10:52.22d-d-cis there any news about the walnut ?
12:37.12*** join/#edev GPSFan (n=kenm@12.10.255.248)
12:37.12*** mode/#edev [+o GPSFan] by ChanServ
12:55.16*** join/#edev redblue (i=star@ppp064.109-253-207.mtl.mt.videotron.ca)
15:09.53*** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca)
16:02.55d-d-cis there any news about the walnut ?
18:10.52*** join/#edev FireEgl (i=Proteus@2001:5c0:84dc:0:211:9ff:feca:b042)
18:48.33d-d-cis there any news about the walnut ?
19:16.18GPSFand-d-c: what news might you be looking for? the DHT-Walnut project is "mature" in that there aren't any more boards available, there is a 2.4 & 2.6 kernel working, there are 2 bootloaders available, and a few projects have been done with the resulting system.
19:21.52*** join/#edev redblue (i=star@ppp061.110-253-207.mtl.mt.videotron.ca)
19:42.38d-d-cGPSFan the walnut is not mature
19:42.56d-d-cit is only able to boot linux, it has a lot of issues with DMA
19:43.10d-d-cit can't play sound with a pci sound card
19:43.38d-d-cand ... boot 1.2 has a problem with pcihost ...
19:45.12d-d-c-
19:45.16d-d-cwhat about few projects ?
19:45.33d-d-cthere is only a not working project on the net
19:45.53GPSFand-d-c: so you say, but that's not my experience. I use ppcboot 1.1.2 not uboot. I play music through it every day using a CS4280 PCI sound card.
19:46.16d-d-cmmmm ppcboot ?
19:46.34d-d-cit may be ppcboot inizialize differently the pci bus
19:46.52d-d-cinteresting ... i'm talking about u-boot v1.1.4 or v1.2
19:47.12d-d-ctoday we understand there is something wrong in the way it inizialize the pci
19:47.33d-d-con the elinuxwiki it is described the uboot ...
19:47.49d-d-c-
19:48.18d-d-cdoes ppcboot need to be patched to boot linux from hd ?
19:48.58GPSFanppcboot was the ancestor of uboot. the original bootloader of the DHT was ppcboot 1.1.2 I ported a later version (1.1.6 I think) and dfarnsworth got uboot running on it.
19:49.06*** join/#edev ico2 (n=ico2@modem-571.cougar.dialup.pol.co.uk)
19:50.06d-d-cso ... i could download the last version of ppcboot, compile it, and it will boot linux from the hd ?
19:50.26d-d-cor i need to patch it as i need to patch ucboot v1.2 ?
19:51.07GPSFanhttp://elinux.org/wiki/DHT-Walnut?action=AttachFile&do=get&target=ppcboot1.1.6.1.bin is the binary of that bootloader from the elinux web site.
19:51.27GPSFanno need to patch & compile it.
19:51.36d-d-cit is the binary
19:51.45d-d-cand i'm talking about the source
19:52.01GPSFanit is not trivial to install though, the wiki has instructions, but without jtag it is possible to brick it.
19:52.27d-d-cwell ok
19:52.29d-d-cbut
19:52.36d-d-ci'd like to recompile
19:52.41GPSFanftp://ftp.denx.de/pub/ppcboot/ppcboot-1.1.6.tar.bz2 source... & patch...http://elinux.org/wiki/DHT-Walnut?action=AttachFile&do=get&target=patch-ppcboot-1.1.6-km2
19:53.02d-d-cthank you
19:53.59GPSFannp, I built my toolchain using buildroot, and could build the kernel as well as userspace with it.
19:54.18d-d-cthe kernel is ok
19:54.24d-d-cthe problem is the bootloader
19:54.29d-d-cas i told before
19:54.42d-d-cuboot doen't inizialize correctly the kernel in pci
19:54.56d-d-c<PROTECTED>
19:55.07d-d-cis a reason of kernel panic, for example
19:55.42GPSFanI have used ati video cards as well as a usb-2 card and 2 kinds of pci sound cards, all work.
19:55.50d-d-cno
19:56.01d-d-cyou are using an other bootloader
19:56.08d-d-ci can assure uboot v1.1.4
19:56.17d-d-chas been tested with matrox and it works
19:56.29d-d-cwhile it sucks with soundboards
19:56.33GPSFanread & understand this: http://elinux.org/wiki/DHT-WalnutInstallingTheUpdatedBootloader before flashing the bootloader. I use ppcboot. I just never got around to trying uboot.
19:56.53d-d-cas i told !
19:57.06d-d-cdifferent inizialization
19:57.21d-d-cand we are working to understand where and what is wrong with uboot
19:58.55d-d-c-
19:59.19d-d-ci will try ppcboot and compare the 2 sources
19:59.27GPSFanI use uboot on another platform (IXP425) and it handles PCI devices ok.
19:59.39d-d-cdamn the hell
19:59.44d-d-cDMA is the problem !
20:00.02d-d-cboot uboot v1.1.6, or v1.2
20:00.06d-d-cand do this
20:00.07d-d-csetenv pciconfighost yes
20:00.10d-d-cboot linux
20:00.19d-d-cand you will see a kernel panic !
20:00.20d-d-ctry !
20:00.36d-d-cthen setenv pciconfighost
20:00.39d-d-creboot linux
20:00.44d-d-cand try to play sound
20:00.54d-d-cyou will hear no sound !
20:02.39GPSFanoops, excuse me that was Redboot on the IXP425. I've got uboot on something, let me poke around a bit... sorry.
20:02.59d-d-c:)
20:08.56GPSFanah, uboot v1.1.4 on http://www.neurosaudio.com/osd/osd.asp. Haven't done much developing on it, and it doesn't have a pci bus , so can't really test any PCI related stuf.
20:09.22d-d-cit has been patched
20:09.30d-d-cadding what is missing
20:09.51d-d-c-
20:10.24lyakhsorry, do I understand it right that there are fundamental doubts in u-boot's ability to initialize / use PCI?
20:10.25d-d-ci need an information: could ppcboot use the setenv of uboot ?
20:10.35d-d-cyes
20:10.53d-d-cuboot doen't inizialize correctly the pci
20:10.57d-d-cas i can demostrate
20:11.05GPSFanI have no direct experience with uboot on anything but the OSD.
20:11.18lyakhok, linkstation / kurobox u-boot 1.1.4 and other newer git versions. eth and IDE on PCI work
20:11.30d-d-cdamn the hell
20:11.38d-d-cIDE/ATA work !
20:11.49d-d-cpci-soundcard no !
20:11.58d-d-cthe bad inizialization is around dma !
20:11.59lyakhotherwise, I am surprised linux depends on boot-loader initializing PCI...
20:12.01d-d-cas i told 3 times
20:12.03GPSFanwhich soundcard?
20:12.17d-d-call soundcard that need DMA
20:12.26GPSFanwhich soundcard?
20:12.36d-d-ctested: soundblaster, ALS400, endoniq1370
20:12.46d-d-cnone of them are working with ALSA
20:12.59d-d-cevery inspections show a problem in DMA
20:13.13lyakhisn't your IDE using DMA?
20:13.22d-d-cno
20:13.25lyakheth?
20:13.33lyakhanything else on PCI?
20:13.41d-d-clyakh: easy experiment, do the following with uboot
20:14.04lyakhhas stefanro given you a patched u-boot version?
20:14.05d-d-csetenv pciconfighost yes
20:14.07d-d-cthen boot
20:14.26d-d-cstefanro is working with me to understand the proble
20:14.35d-d-chowever, do the setenv pciconfighost yes
20:14.51d-d-cand you will see by yourself
20:14.51lyakhok, wait a sec
20:14.56lyakhstefanro WILL fix it, don't worry
20:15.00d-d-cyou will see linux not able to work
20:15.05d-d-cthen reset to uboot
20:15.07d-d-cand type
20:15.08d-d-csetenv pciconfighost
20:15.15d-d-cthis will clean the var
20:15.24d-d-creboot and you will see linux well booting
20:15.30d-d-c-
20:15.43d-d-cthen have a look into includes/walnut.h
20:15.55d-d-cand you will see pci-host not enabled
20:16.16d-d-c-
20:16.25d-d-cconclusion: something wrong in uboot
20:16.45d-d-c-
20:16.47d-d-cdo pci long
20:16.57d-d-cand you will see no bar addressed
20:18.04d-d-cthis should confirm ...
20:18.34lyakhok, just set "setenv pciconfighost yes" and booted fine
20:18.46d-d-cimpossible
20:18.49lyakhwhat does it do?
20:18.54d-d-cuboot v1.1.4 ?
20:19.11lyakhU-Boot 1.1.4-gf7b190d3-dirty (Nov  6 2006 - 19:17:58)
20:19.17d-d-cit should stop the boot with an error in dma
20:19.28d-d-cthe original one
20:19.29d-d-cok
20:19.37d-d-cyou are not able to play sound at all
20:19.38d-d-ctry !
20:19.56lyakhwell, I really think the kernel should initialize stuff itself
20:20.00lyakhsorry, no soundcard
20:20.07d-d-cit doesn't
20:20.19d-d-cwe thought too
20:20.22d-d-cwe are wrong
20:20.25lyakheth works
20:20.30d-d-cdam i know !
20:20.35d-d-ceth and ata work
20:20.39d-d-cthe sound doesn't !
20:21.01*** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca)
20:21.02lyakhyou said no PCI devices work that use DMA, I think, my eth uses DMA
20:21.28d-d-cDMA has a special features in the kernel
20:21.32d-d-cif you enabled it
20:21.39lyakh?
20:21.40d-d-cEVERY DEVICES STOP TO WORK
20:21.47d-d-chowever, the sound doen't work
20:21.57GPSFand-d-c: calm down.
20:22.00lyakhwhat special feature?
20:22.10d-d-c:'(
20:22.20d-d-ci hate to reapet thing 1000 times
20:22.58lyakhsure, don't
20:24.01d-d-ci go testing the ppcboot
20:24.01d-d-cbbl
20:24.01GPSFando you have IBM4xx DMA support turned on in the kernel?
20:24.01d-d-cthe sound card need it
20:24.01GPSFanyes
20:24.01d-d-cthe problem is in it
20:35.56*** join/#edev dijenerate (n=dijenera@69.73.255.253)
20:42.02*** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca)
20:58.34*** join/#edev Heet (n=lucas@a82-92-5-118.adsl.xs4all.nl)
20:58.36agaffneywow, that conversation was looking a bit one-sided
21:31.46*** join/#edev ico2_ (n=ico2@modem-164.chameleon.dialup.pol.co.uk)
23:48.48*** join/#edev ico2 (n=ico2@modem-3746.cougar.dialup.pol.co.uk)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.